INDEX

Explanations

mentions of notable impacts or changes

oai_token-act-pair · gpt-3.5-turbo

highly emphasized or impactful contexts in discussions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

atures

-0.92

hops

-0.89

agues

-0.88

oots

-0.88

classes

-0.82

styles

-0.78

endars

-0.78

chairs

-0.78

events

-0.78

okes

-0.77

POSITIVE LOGITS

 inability

1.43

 ability

1.31

 fact

1.30

 insistence

1.25

 willingness

1.25

 absence

1.21

 unwillingness

1.20

 notion

1.18

 realization

1.17

 tendency

1.16

Activations Density 0.196%

mentions of notable impacts or changes

highly emphasized or impactful contexts in discussions

No Comments

No Known Activations